Particle atomic layer deposition
- 450 Downloads
The functionalization of fine primary particles by atomic layer deposition (particle ALD) provides for nearly perfect nanothick films to be deposited conformally on both external and internal particle surfaces, including nanoparticle surfaces. Film thickness is easily controlled from several angstroms to nanometers by the number of self-limiting surface reactions that are carried out sequentially. Films can be continuous or semi-continuous. This review starts with a short early history of particle ALD. The discussion includes agitated reactor processing, both atomic and molecular layer deposition (MLD), coating of both inorganic and polymer particles, nanoparticles, and nanotubes. A number of applications are presented, and a path forward, including likely near-term commercial products, is given.
KeywordsAtomic layer deposition Particle ALD Nanoparticle Nanolayers Coating
Generically, a CVD reaction can be divided into successive surface reactions that occur solely on a particle surface, reacting surface functional groups, to define ALD. For example, to deposit aluminum oxide (Al2O3), the binary CVD reaction between trimethylaluminum (Al(CH3)3) (TMA) and water vapor (H2O) generates methane (CH4) as a byproduct according to Reaction (1).
For ALD, the binary CVD Reaction (1) can be split into two successive surface reactions, shown in the subsequent texts. The asterisk (*) denotes a surface reaction, and so, if TMA and H2O are not present simultaneously (as is done in CVD), the reaction occurs entirely at the surface and no gas-phase reaction-producing nanoparticles occur. The sequential surface reactions A (2) and B (3) are then repeated (cycled) in order to grow an ultrathin and conformal film (Ott et al. 1997a, b) (Fig. 3c, d):
Deposition is controlled at the atomic level by self-limiting surface reactions (George et al. 1996). The process is independent of line of sight. Hence, uniform and conformal deposition will occur on high-aspect ratio porous structures or on particles in particle beds because the surface chemistry is self-passivating (Ott et al. 1997a, b). Precursors do not self-react; they only react with the functionalized surface produced by the reaction with the complementary precursor. Once the reaction is completed at one surface site, the reactants will continue to travel down the high-aspect ratio pore or convoluted path in the particle bed and reach the unreacted surface sites. Consequently, the deposition produced by each surface reaction only proceeds until no further active sites are accessible to the precursor on the substrate surface, making the deposition self-limiting. The thickness of the film is only dependent on the number of times the surface reactions are cycled, i.e., AB cycles. A review of ALD chemistry is given by George et al. (2000) and George (2010).
Early history of particle ALD
ALD was pioneered in Finland in the early 1970s (Suntola and Jorma 1977). Particle ALD was pioneered in the late 1990s at the University of Colorado (George et al. 2003, 2004, 2005), and it is characterized by the ability to coat primary particles (not agglomerating them in the process), including nanoparticles with conformal films. The initial objectives for particle ALD were to deposit pinhole-free films as thin as possible in order to provide an environmental barrier coating (EBC) or to functionalize the particle surface for a specific application. A major effort was identifying a minimum film thickness for a barrier or to achieve a particular effect.
Although an early focus for ALD on particles was for complete pinhole-free passivating films, research was also directed towards providing ALD films without complete surface coverage in order to best maintain substrate properties. For example, boron nitride (BN) has a high thermal conductivity and its addition to composite materials is important for enhanced thermal management applications. In particular, the miniaturization of microelectronic devices has led to larger heat dissipation that requires higher thermal conductivity packaging materials. One impediment to the addition of BN particles in composites is the inertness of the BN surface basal planes. The unreactive BN surface limits the coupling between the BN particles and the epoxy matrix and lowers the BN particle loading. Hence, ultrathin films, or partial films (i.e., non-uniform, semi-continuous films), are needed to alter the chemical activity of the BN surface without significantly degrading the thermal conductivity of the BN particles.
Agitated particle bed reactors
Coating primary nanoparticles
Even though the nanoparticles were fluidized with larger aggregates (Fig. 12a), they were individually coated with conformal films (Fig. 12b). The particles were not agglomerated. Part of the explanation is proposed by Hakim et al. (2005c) who investigated the fluidization of nanopowders using a high-speed laser imaging system in real time. Although fluidization of aggregates is dictated by interparticle forces, they found that fluidized aggregates show a dynamic behavior where outer edges are shed and picked up by other aggregates. The relatively large size of aggregates of nanoparticles and their frequent collisions with other large aggregates while in continuous flow promote this dynamic behavior. So, during fluidization, aggregates of nanoparticles continuously break apart and form. The aggregates do not maintain a stagnant size or shape. This “dynamic equilibrium,” or, dynamic aggregation, between inertial and cohesive forces is a unique characteristic of fluidized nanoparticles. In this manner, all particle surfaces are exposed to the surrounding gas and ALD can deposit conformal films on the agitated primary nanoparticles.
Molecular layer deposition
Ultrathin microporous/mesoporous metal oxide films
Coating polymer particles
Improving polymer properties can benefit the multitude of uses for polymers. The high gas permeability of polymers is one property that limits their use in various food, medical, and electronic packaging applications (Chatham 1996; Erlat et al. 1999; Weaver et al. 2002). Inorganic materials typically have a much lower gas permeability than polymers. When used as coatings on polymers, these inorganic materials can serve as gas diffusion barriers and can dramatically improve the polymer performance (Chatham 1996; Erlat et al. 1999; Weaver et al. 2002). However, polymers are thermally fragile. Low-temperature deposition techniques, such as sputtering, evaporation, and plasma-enhanced chemical vapor deposition (CVD), have been required to deposit the inorganic diffusion barrier (Erlat et al. 1999). Because inorganic materials are brittle, thin inorganic diffusion barriers on polymers are needed to maintain polymer flexibility without cracking. The optimum thickness for maximum flexibility is as thin as possible, but thick enough to provide the specific barrier performance. For these small thicknesses, line-of-sight deposition techniques, such as sputtering and evaporation, are limited by defects and pinholes. The continuous and pinhole-free ALD film characteristics are important for gas-diffusion barriers.
Porous polymer/ceramic composite materials and use as templates
Porous alumina and other ceramic particles or structures with crystallized frameworks and controlled nanometer wall thickness can be easily fabricated by ALD. A sacrificial template, such as a polymer, can be coated by ALD, and then, the sacrificial substrate was removed leaving behind a unique ceramic structure. Liang et al. (2012b) demonstrated that the highly porous ALD-coated PS-DVB polymer particles (Liang et al. 2007a) could be calcined in air to remove the polymer template and leave a porous ceramic structure with precise wall thickness corresponding to the ALD growth rate and number of ALD cycles carried out. Surprisingly, for Al2O3 ALD, the structure did not collapse and mimicked the starting morphology of the polymer. A mesoporous structure of crystalline Al2O3 with a high specific surface area and large pore volume was formed for calcination temperatures above 600 °C. Porous crystalline alumina with a surface area of 80–100 m2/g was obtained and was thermally stable at 800 °C. Such porous alumina particles may find wide application in nanotechnology and catalysis.
Metal ALD films and seed layers on polymer particles
Tungsten (W) ALD was investigated on a variety of polymer particles, including polyethylene (PE) MW-1100, polyvinylchloride (PVC) MW-90,000, polystyrene (PS) MW-190,000, and polymethylmethacrylate (PMMA) MW-15,000 (Wilson et al. 2008). The polymer particles were placed in a rotary ALD reactor without any prior treatment. The W ALD was performed at 80 °C using tungsten hexafluoride (WF6) and disilane (Si2H6) as the gas phase reactants. The nucleation of W ALD directly on the polymer particles at 80 °C required > 50 AB cycles. In contrast, the polymer particles treated with only 5 AB cycles of Al2O3 ALD to provide a “seed layer” were observed to blacken after 25 AB cycles of W ALD. XPS analysis of the W 4f peaks after W ALD on the polymer particles was consistent with a WO3 thickness of 29 Å covering the W ALD film. The oxidation of the W ALD film may be dependent on the radius of curvature of the polymer particles. W ALD on polymers may have applications for flexible optical mirrors, electromagnetic interference shielding, and gas diffusion barriers.
Functionalized particulate materials
EBCs for passivation
Recently, Hoskins et al. (2018) have shown that an Al2O3 or mullite (3Al2O3:2SiO2) coating on SiC can reduce steam oxidation of the SiC at 1000 °C (20 h) by up to 62% for a 10-nm-thick ALD film on micron-sized SiC particles. These results are comparable to CVD films that are three orders of magnitude thicker and support the conclusion that the superior ECB properties using ALD films is the result of the films being free of pin-holes. The mullite ECB is preferred since it has a thermal expansion coefficient similar to SiC. These results have substantial potential for ALD ECB coatings for passivation of SiC microchannel heat-exchanger surfaces.
Particle ALD ECBs may also mitigate problems with excess helium in spent nuclear fuel (Zhang et al. 2018). Helium gas accumulation from alpha decay during extended storage of spent fuel has potential to compromise the structural integrity the fuel. Zhang et al. (2018) reported results obtained with surrogate nickel particles which suggests that alumina formed by ALD can serve as a low-volume fraction, uniformly distributed phase for retention of helium generated in fuel particles, such as uranium oxide. Thin alumina layers may also form transport paths for helium in the fuel rod, which would otherwise be impermeable. Micron-scale nickel particles, representative of uranium oxide particles in their low helium solubility and compatibility with the alumina synthesis process, were homogeneously coated with alumina approximately 3–20 nm by particle ALD using a fluidized bed reactor. Particles were then loaded with helium at 800 °C in a tube furnace. Subsequent helium spectroscopy measurements showed that the alumina phase, or more likely a related nickel/alumina interface structure, retained helium at a density of at least 1017 atoms/cm3. High-resolution transmission electron microscopy (HRTEM) revealed that the thermal treatment increased the alumina thickness and generated additional porosity. Results from Monte Carlo simulations on amorphous alumina predicted that the helium retention concentration at room temperature could reach 1021 atoms/cm3 at 400 MPa, a pressure predicted by others to be developed in uranium oxide without an alumina secondary phase. This concentration is sufficient to eliminate bubble formation in the nuclear fuel for long-term storage scenarios, for example.
Particle ALD can modify rheological behavior of nanoparticle suspensions (Hakim et al. 2007b) and of slurries and bulk powders comprised of 1 to 5 μm diameter particles (Kilbury et al. 2012). Microfine zinc powders, similar to those used in alkaline batteries, have been coated using boron nitride (BN) ALD films of sub-nanometer thickness or about 0.1 wt.%. The low-surface energy coatings reduced the cohesion of 1–5 μm particles by 52%. A highly loaded slurry of the same material in concentrated KOH showed a 10–30% reduction in slurry viscosity over a range of shear rates, with a shear thinning effect at high shear rates. Boron nitride (BN) platelets were coated using Al2O3 and SiO2 films to change the surface properties from hydrophobic to hydrophilic. As noted previously by Ferguson et al., the platelet structure of the BN provided for reactive surface functional groups on the edges while the basal planes only had an electron pair associated with nitrogen. Hence, while it was possible to coat the entire BN particle with Al2O3, a SiO2 coating was “patchy” and primarily on reactive edges. The coated and uncoated powders were dispersed into an epoxy to evaluate the solids loading to viscosity ratio. The ALD films improved the particle–resin adhesion and decreased the viscosity of an equivalently loaded slurry of uncoated powder (Fig. 8). Viscosity was reduced the most when the entire particle surface was coated by either Al2O3 or a SiO2/Al2O3 composite film. Coated microfine nickel, aluminum, and iron powders were also dispersed into epoxies, and lower viscosities and yield stresses were observed due to ceramic–epoxy interactions being more favorable than metallic–epoxy interactions.
Zinc oxide (ZnO) and TiO2 are wide (~ 3.3 eV) and medium (~ 3.0–3.2 eV) bandgap semi-conductor materials, respectively. They find use in a variety of optical, optoelectronic, and piezoelectronic applications, as well as in commodity markets, such as pigments, sunscreens, cosmetics, and even food products. TiO2 is also a well-known photocatalyst with a large propensity to photodegrade surrounding media because of free-radical generation in the presence of UV-light irradiation. Particle ALD applications using the UV-absorbing properties of ZnO and TiO2 include UV-driven water purification (King et al. 2009c; Zhou et al. 2010) and sunscreen/personal care (King et al. 2008b, c, d, e) products.
The benefits of surface photocatalysts can be integrated with known magnetic separation techniques by creating photoactive magnetic particles (Zhou et al. 2010). Iron-based magnetic nanoparticles were produced by decomposition of iron oxalate powder, and then, a titanium dioxide (TiO2) thin film was deposited on the synthesized iron nanoparticles with an in situ atomic layer deposition (ALD) process at 100 °C using TiCl4 and H2O2 as precursors. However, because of the high surface area, the iron nanoparticles were unstable and spontaneously oxidized when exposed to H2O2 during the TiO2 ALD process, thus reducing the magnetic moment of the core particles. As an improvement in the process, prior to the TiO2 deposition, an aluminum nitride (AlN) film was deposited in situ to coat and passivate the iron core particles. The AlN ALD was performed at 250 °C with trimethylaluminum (TMA) and ammonia (NH3) as precursors. This passivation provided a significant decrease in the iron oxidation as determined by X-ray diffraction and magnetization measurements. Photoactivity of the TiO2 film was demonstrated by decomposition of methylene blue solution under ultraviolet irradiation.
Many catalytic processes, such as catalytic combustion, steam reforming, and automobile exhaust control, have reaction temperatures typically in excess of 300 °C. Metals are dispersed on high-surface area supports so that the resulting metal nanoparticles have a high fraction of their atoms on the surface. Catalysts can be designed using particle ALD (O’Neill et al. 2015). However, supported metal catalysts deactivate at high temperatures when these metal particles sinter to form larger particles. Particle ALD is used to reduce sintering (Feng et al. 2011; Liang et al. 2011; Lu et al. 2012; Kim et al. 2018; Phaahlamohlaka et al. 2018).
Enhanced thermite materials
Ultrafast metal-insulator varistors
Solar thermochemical water and CO2 splitting active materials
Li-ion battery materials
ALD’s applications have been extended to sodium-ion (Meng 2017a, b) batteries and lithium-sulfur (Sun et al. 2018) batteries. It is anticipated that particle ALD will play a key role in the low-cost coating of LIB cathode materials since preventing capacity fade is a key element to LIBs gaining widespread acceptance.
EL Phosphor CVD vs. ALD Barrier Layer Performance
300 Al2O3 cycles
600 Al2O3 cycles
Initial brightness (%)
24 h (%)
100 h (%)
Maint. (%, 100 h)
Coating thickness (Å)
Ceramic particle sintering aid additives
Perspectives, challenges, and path forward
Particle ALD allows the design and fabrication of complex atomic nanostructures using particulate precursors. Primary particles, including nanoparticles and high-aspect ratio nanotubes, can be coated if one uses an agitated processing system. Film thickness can vary for different applications and ranges from sub-nanometer to tens of nanometers thick. Films can be uniform or non-uniform depending upon the functionalization of the particle surfaces and the nucleation required for a given ALD chemistry. Particle ALD is a low-cost process due to the ability to use almost 100% of sequential precursors (Fig. 9). This recognition of low cost was a major stumbling block in the consideration and adaptation of particle ALD for commercial applications. Another major hurdle was convincing possible users that primary (individual) particles, including nanoparticles, could be coated without agglomeration (Hakim et al. 2005a, b). A major opportunity exists for developing continuous low-cost spatial particle ALD processes for huge tonnages of commercial products that require only a few ALD cycles, such as LIB materials. Low cost will require those continuous processes to efficiently use expensive precursors with near 100% usage similar to the efficiency of batch-fluidized beds. Further, because of the handling of such large quantities of fine powders reacting with sequential gases, those spatial ALD processes need to be relatively simple and to avoid being “solids processing nightmares” having major powder handling issues. While most applications have yet to be discovered, likely initial commercial products will employ particle ALD for cathode battery and lighting application material passivation, as well as catalyst sintering prevention and as sintering additives for advanced ceramic materials. Annual citations for particle ALD have grown exponentially with time from one citation in 1999 to 3933 citations in 2017, totaling more than 21,500 citations according to the Web of Science. Exponential interest in particle ALD is continuing.
Compliance with ethical standards
Conflict of interest
A.W. Weimer has a significant financial interest in ALD NanoSolutions.
- George SM, Ferguson JD, Klaus JW (2000) Atomic layer deposition of thin films using sequential surface reactions. In: Wadley HNG, Gilmer GH, Barker WG (eds) New methods, mechanisms and models of vapor deposition, vol 616, pp 93–101Google Scholar
- George SM, Ferguson JD, Weimer AW (2003) Atomic layer controlled deposition on particle surfaces. (U.S. Patent 6,613,383)Google Scholar
- George SM, Ferguson JD, Weimer AW, Wank JR (2004) Insulating and functionalizing fine metal-containing particles with conformal ultra-thin films. (U.S. Patent 6,713,177)Google Scholar
- George SM, Ferguson JD, Weimer AW, Wank JR (2005) Nanocoated primary particles and method for their manufacture. (US. Patent 6,913,827)Google Scholar
- Kim IS, Li ZY, Zheng J, Platero-Prats AE, Mavrandonakis A, Pellizzeri S, Ferrandon M, Vjunov A, Gallington LC, Webber TE, Vermeulen NA, Penn RL, Getman RB, Cramer CJ, Chapman KW, Camaioni DM, Fulton JL, Lercher JA, Farha OK, Hupp JT, Martinson ABF (2018) Sinter-resistant platinum catalyst supported by metal-organic framework. Angew Chem Int Ed 57(4):909–913CrossRefGoogle Scholar
- O’Toole R, Bartel C, Kodas M, Horell A, Ricote S, Sullivan N, Gump C, Musgrave C, Weimer A (2018) Particle atomic layer deposition of alumina for sintreing yttria-stabilized cubic zirconia. J Am Ceram Soc. https://doi.org/10.1111/jace.16091
- Patel RL, Xie H, Park J, Asl HY, Choudhury A, Liang XH (2016a) Significant capacity and cycle-life improvement of lithium-ion batteries through ultrathin conductive film stabilized cathode particles. Adv Mater Interfaces 3(13):1500046Google Scholar
- Patel RL, Jiang YB, Choudhury A, Liang XH (2016b) Employing synergetic effect of doping and thin film coating to boost the performance of lithium-ion battery cathode particles. Sci Rep 6:25293Google Scholar
- Snyder MQ, Trebukhova,SA, Ravdel B, Wheeler MC, DiCarlo J, Tripp, CP, DeSisto WJ (2007) Synthesis and characterization of atomic layer deposited titanium nitride thin films on lithium titanate spinel powder as a lithium-ion battery anode. Journal of Power Sources 165:379–385Google Scholar
- Suntola T, Jorma A (1977) Method for producing compound thin films. (U.S. Patent 4,058,430)Google Scholar
- Wank J, Buechler K, Hakim L, George S, Weimer A (2004a) Coating fine iron particles with an oxidation-resistant gamma alumina nanolayer using ALD in a fluidized bed reactor. In: Fluidization XI—present and future of fluidization engineering. ECI Intl, Brooklyn, pp 603–610Google Scholar
- Weaver MS, Michalski LA, Rajan K, Rothman MA, Silvernail JA, Brown JJ, Burrows PE, Graff GL, Gross ME, Martin PM, Hall M, Mast E, Bonham C, Bennett W, Zumhoff M (2002) Organic light-emitting devices with extended operating lifetimes on plastic substrates. Appl Phys Lett 81(16):2929–2931CrossRefGoogle Scholar
- Weimer A, George S, Buechler K, Spencer A, McCormick J (2014) Methods for producing coated phosphors and host material particles using atomic layer deposition methods. US Patent 8,637,156)Google Scholar
- Yang F, Brede J, Ablat H, Abadia M, Zhang LB, Rogero C, Elliott SD, Knez M (2017) Reversible and irreversible reactions of trimethylaluminum with common organic functional groups as a model for molecular layer deposition and vapor phase infiltration. Adv Mater Interfaces 4(18):1700237Google Scholar
- Zhang SL, Yu E, Gates S, Cassata WS, Makel J, Thron AM, Bartel C, Weimer AW, Faller R, Stroeve P, Tringe JW (2018) Helium interactions with alumina formed by atomic layer deposition show potential for mitigating problems with excess helium in spent nuclear fuel. J Nucl Mater 499:301–311CrossRefGoogle Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.